Routing decisions made by unmanned aerial vehicle (UAV) swarms are affected by complex and dynamically changing topologies. A centralized routing algorithm imposes the entire computational burden on one module, and the high data dimensionality renders computation burdensome. In this letter, we develop a multi-agent reinforcement learning-based routing algorithm for a UAV swarm. The UAVs are trained in a data-driven manner to make distributed routing decisions. Factors that include channel quality, UAV movement, UAV overhead, and the extent of neighbor variation are incorporated into link quality assessment. Long short-term memory is used to improve the Actor and Critic networks, and more information on temporal continuity is added to facilitate adaptation to the dynamically changing environment. Simulations show that the proposed routing algorithm reduces data transmission delay and enhances the transmission rate compared with traditional routing algorithms.

In this paper, a downlink security-aware resource allocation problem with delay constraint via spectrum sensing is modeled as a mixed integer non-linear problem for non-orthogonal multiple access-based cognitive radio network. The security-aware resource allocation is subject to constraints in required delay for each secondary user, maximum number of accessed secondary users at each subchannel, total interference power threshold introduced to primary users, and total power consumption at secondary BS. The security-aware resource allocation is based on channel state information at the physical layer and queue state information at the link layer. According to the queue buffer occupancy, a probability upper bound of exceeding the maximum packet delay based on M/D/1 queuing model is analyzed in terms of a required minimum secrecy transmission rate. Then, a secondary user scheduling problem and a power allocation problem are solved, separately. Finally, the secondary user scheduling problem is solved via greedy algorithm, and a power allocation algorithm is proposed by successive convex approximation method. The simulation results demonstrate that the performance of proposed algorithms can be improved significantly.

Non-orthogonal multiple access (NOMA)-based cognitive wireless networks can improve the spectral efficiency to utilize the vacant spectrum resource and exploit the power domain diversity. In this work, we formulate a max-min resource allocation problem for video traffic in NOMA-based cognitive wireless networks as a mixed integer non-linear programming (MINLP) problem. The max-min video transmission problem is subject to the constraints of maximum accessed user number at each subchannel, total available energy of each secondary user, video encoding characteristics, and interference power threshold. To solve the formulated MINLP problem, we divide it into two subproblems, i.e., a power allocation and secondary user scheduling subproblem, and a video packet scheduling subproblem. Firstly, we apply successive convex approximation to transform the joint power allocation and secondary user scheduling subproblem into a bi-convex programming problem. Secondly, binary search and dual decomposition methods are combined to obtain the approximated optimal power allocation and secondary user scheduling solutions. Finally, we propose a heuristic packet scheduling algorithm. Simulation numerical results show that the proposed algorithm improves the video quality and guarantees the fairness among different secondary users.

In this work, an uplink secondary Internet of Things (IoT) device scheduling and power allocation problem based on imperfect channel state information (CSI) and imperfect spectrum sensing is investigated for industrial cognitive IoT over cognitive heterogeneous non-orthogonal multiple access (NOMA) networks. The joint secondary IoT device scheduling and power allocation problem maximizes the network throughput subject to total power constraint at each secondary IoT device, proportional fairness transmission rate among different secondary IoT devices, maximum number of accessed secondary IoT devices for each subchannel, and interference power threshold constraint at each primary base station (BS). Firstly, successive convex approximation method is adopted to transform the original resource allocation problem into a bi-convex programming problem. Then, we employ the dual decomposition method to analyze the secondary IoT device scheduling criterion and obtain the power allocation close-form expression. Finally, a joint power allocation and secondary IoT device scheduling algorithm with the proportional fairness criterion is proposed. Numerical simulation results demonstrate that the throughput and fairness for the proposed algorithm are better than that of other resource allocation algorithm significantly.

Routing protocols do not respond quickly to environmental changes due to the high mobility of nodes in the Flying Ad Hoc Network (FANET), to obtain reliable transmission links. This paper proposes an adaptive link maintenance method based on deep reinforcement learning (DRL-MLsA), which can dynamically adjust the time interval of broadcasting Hello packets. This method can cope with the highly dynamic network environment, and adapt to both active routing and table-driven routing protocols. The method considers the channel model of the signal and investigates the impact of UAV communication range on link maintenance. We can get an agent by perceiving the degree of changes in the number of neighbors in a dynamic environment. The optimal broadcast cycle was obtained to maximize the energy of the node to send and receive task data. We substituted the single-output network model with a competitive network to overcome the reward overestimation problem, which also improves the convergence speed of the algorithm. Simulation results showed that DRL-MLsA can reduce the communication overhead for link maintenance, while at the same time increase the throughput of the network and decrease the packet loss of transmission.